Section III

CASE HISTORY: THE

Session 22

Chairman: KenBiglane Environmental Protection Agency

Co-Chairman: Wilmot Hess National Oceanic and Atmospheric Administration

THE AMOCO CADIZ CLEANUP OPERATIONS- AN OVERVIEW OF THE ORGANIZATION, CONTROL, AND EVALUATION OF THE CLEANUP TECHNIQUES EMPLOYED

Pierre Bellier (CEDRE) and Georges Massart (CEDRE-CNEXO) B. P. 337 29273 Brest Cedex France

ABSTRACT: The organization used for work on the Amoco Cadiz were as harmless as possible. All told, less than 3,000 tons of disper- oil spill, and the changes in this organization required by the size of sants was used; some chalk, used as a sinking agent, was also em- the disaster, are explained and details are provided on the techniques ployed but only to protect the Bay of Brest which was threatened by used in the cleanup. These techniques had as their goals: protection of some oil patches a few days after the spill. selected areas; pumping of oil where possible; and cleanup of the The onshore operations organizations for the Departments of Finis- beaches, the shingle shores, rocky areas, and harbors, and disposal of tere and Cötes du Nord were different because the responsibility came the oily debris. The spill involved some 223,000 tons of oil spilled under the "Prefet" (the Government's representative) in each depart- along 400 kilometers of coast. Almost 10,000 persons worked on the ment. After a few days of reflection, hesitation, and improvising, the project during the busiest period. As a result, almost 200,000 tons of Finistfcre Department became organized in this way: a managing staff oil and debris were pumped and gathered. However, less than 20,000 set up in Ploudalmézeau, a small town near Portsall, while a technical tons of oil was finally retrieved after separation from the total mass team worked in Brest. This team gathered people belonging to the of material obtained from the coastal zone. Ministry of Equipment, to the Institut Francais du Petrole (IFP, the French Institute) and to the Centre Oceanologique de Bre- tagne (Research Center of the French National Agency for Ocean Ex- ploitation). This team tried to find appropriate answers to the numer- ous questions coming in from the people working on the coast who The cleanup organization faced difficulties and problems never encountered before. The prevailing winds for a month after the disaster were from the When the A moco Cadiz went on the rocks on March 16, 1978 (Fig- southwest to the northwest (except for the two first days after the acci- ure 1), it soon became evident that her 223,000 tons of light crude oil dent during which it had been blowing from the north) and spread the would spoil a large part of the coast. What administrations oil to the east of Portsall. Less than one week after the accident, the were concerned and on what organization could the French authorities oil reached the Cötes du Nord Department area which, of course, had rely to coordinate and manage all the cleanup operations? been forewarned. In this department, the headquarters were in Lan- Based on previous experiences which we had to deal with in France nion and included both technical and operational teams, thereby (Torrey Canyon, Olympic Bravery, and Boehlen having been the most allowing these specialties to work more closely than in the Finistlre important), it had been previously decided in the "Polmar Plan" that Department. However, even though, the solutions found and tech- the French Navy would be responsible for the fight against pollution niques used differed in some details, cleanup operations between the at sea, and Civil Safety Services would be responsible on shore. How- two departments were accomplished by a small staff at Rennes. This ever, a few days after the grounding of the Amoco Cadiz, it became unit provided and distributed wagons and trucks used to evacuate the obvious that the onshore cleanup operations would require a great wastes from the theater of operations, and provided all types of items deal of people from different administrations, particularly from the which had to be shared between Finistfcre and Cötes du Nord as neces- Army. An overall national coordinator was therefore needed for the sary. operations, which were conducted in the following way: (a) The Navy kept the responsibility for all the offshore operations; and (b) A Civil Safety Representative had the responsibility for all the on- shore operations which utilized firemen from the Civil Safety Attempts to minimize the spill and protect the shore Organization, civil officers from the Ministry of Equipment who have, in France, the responsibility for harbors and other public The first ideas entertained immediately after the accident were to works along the coast, and different Army corps (military engi- burn the oil, to lighten the wrecked ship, and to protect the most vul- neering, communications, transportation, and, of course, infan- nerable parts of the shore. Burning the oil seemed too dangerous, be- try). cause the smoke and unburned oil carried by the smoke would have Weather conditions did not permit use of one or more of the numer- polluted a large inland area. A lot of oil would not have burned and ous kinds of devices which have been successfully used elsewhere in would have remained to pollute the sea. In addition, the experts the open sea. If the weather had been more favorable, the Navy would thought that the part of the oil which would burn would eventually have decided which of these to use. As it turned out, adverse meteoro- evaporate anyway. Lightening the ship by pumping the oil ashore was logical conditions and weathering of the oil reduced these options very not feasible because the 2.5 kilometers (1.5 mile) pipe needed for this quickly to the use of only a few types of dispersante complying with would be almost impossible to lay down due to the roughness of the the recommendations of the Ministry of Environment. The Institut sea. Pumping the oil into a smaller would have been difficult, Scientifique et Technique des Peches Maritimes (ISTPM or Scientific dangerous, and a very long process. In fact, the Amoco Cadiz was and Technical institute for Marine Fisheries) had conducted disper- completely broken up and almost empty before lightening operations sant toxicity tests and was able to indicate those dispersants which could have been attempted eight days later.

141 142 1979 OIL SPILL CONFERENCE

Lee Sept. lies

Brehat lie Grande *>^*a, Roscoff Cfcerros-Guirecft Paumpol

lie d'Ouessant ft

Pte. de St. Mathieu \

0 10 20 30 40 50 60 70 80

Figure 1. Chart showing place names associated with the^moco Cadiz oil spill of March 1978; heavy offshore lines indicate the approximate distribution of beached oil

The most vulnerable parts of the shore were the places where Sufficient boom was not available, and the time left for deploying and were grown, especially in the rivers called Aber Bencít and Aber mooring before the oil arrived would have been too short. In addition, Wrach, the Carantec River and the Morlaix River. These places also by so using the booms, only a few places could have been protected, seemed, to many people, the easiest to protect. In fact, booming thereby making it difficult to explain to the people concerned and re- across these rivers proved useless because of the speed of the tidal cur- quiring a political decision for which the public was not prepared. rents. The whole Bay of Morlaix was protected by a 4 kilometers (2.5 miles) long boom, the maintenance of which required a number of persons under almost permanent alert. This action was considered as Pumping operations successful. It should be noted that the Bay of Morlaix was rather shel- tered, both from severe weather conditions and from the arrival of The other obvious idea was to try to pump as much oil as possible excessive quantities of oil. The boom was laid far enough from the from the shore. Some skimmers were used in the harbors and pro- mouth of the river so as not to be exposed to strong currents and tected areas (Figure 2). Whatever type they were (and a lot of different therefore work efficiently. kinds of devices were used), their efficiency was limited because of the This situation did not exist across the Aber Benoít and Aber Wrach seaweed which blocked the pumps and hoses. Layers of weed two Rivers, where the booms that were deployed were very quickly broken inches thick or more were observed; in fact, in some places pumping up by strong winds and waves during the two weeks after the accident. was possible only with vacuum devices. Vacuum cleaning tanks came It would certainly have been possible to protect some places by using from everywhere in France and even from Belgium and The Nether- the booms in the deflector mode, but this would have required a very lands. These trucks did a good job, but were limited by their weight large amount of boom to protect all the places needing protection. which obliged them to work from piers, boat slips, or roads (Figure 3). CASE HISTORY — THE AMOCO CADIZ 143

Figure 2. ACME skimmers being used in Portsall Harbor dur- Figure 4. Pumping the oil trapped in a small beach pond by ing the first few days after the spill incident use of a "honey wagon's" vacuum pump

Figure 3. Pumping oil from the surface of the water using vac- Figure 5. Temporary storage of oily residue in a lined dug pit uum trucks near Portsall. near Roscoff

Often, when a pumping possibility existed, that is, when the oil was after the accident, the surface oil layers became very thin and dis- gathered in a bay, access was rather difficult. The farmers, driving persed. The emulsified oil was so viscous in the very few places where their tractors pulling vacuum tanks designed to handle liquid manure it remained in heavy concentrations that vacuum trucks could no (called "honey wagons" in the U.S.A.) would come and pump out the longer pick it up. All in all, the total quantity of pumped liquid almost "chocolate mousse" water-in-oil emulsion which spoiled every place reached one hundred thousand tons. In all this product, less than along approximately 400 kilometers (250 miles) of the Brittany coast twenty thousand tons of oil were to be found after treatment in refin- (Figure 4). ing plants. These vacuum tank trailers pumped everything—oil, water, and seaweed. The water was separated as much as possible in the vacuum tanks; the tanks were then emptied through filter-buckets into interim Removal of oil and oily debris stranded along the shore storage of 10-to-20 cubic meter capacity. Regular vacuum trucks then transferred the emulsified oil from these storage tanks into large tank The last stage of the cleanup operation could begin in an efficient trucks, railway tank cars or even into small tanker ships moored in way once there was no longer any oil floating on the sea in appreciable Roscoff harbor for further transport to refining plants. To make the quantity. The risk of further pollution in these areas was rather small, oil transfer easier, some chemical products were used to break the even if high spring tides were able to wash ashore some oil from the emulsions. Good results were achieved, but it was difficult to make still-polluted rocky areas. In any event, the final phase of cleanup people realize that they must add the product either in the "honey operations could not be delayed, as these operations had to be com- wagons" before pumping, or in the interim storage tank whenever a pleted in the tourist areas before summer. vacuum tank was emptied into the storage tank. In some places, where There were four main types of shorelines, each presenting different intermediate storage capacity was lacking, we dug large holes in the problems, which had to be cleaned, namely: sandy beaches, shingle earth which were lined with plastic sheets in order to be oil tight (Fig- strands, rocky areas and stone or concrete construction in harbors, ure 5). This type of work extended for about one month, during which and muddy or marshy areas. up to 300 pumping machines, 150 road trucks and more than 1,500 In removing the oil from beaches, the first operation was gathering persons were engaged in the pumping operations. About one month and removing the stranded oily seaweed. In part, this was done man- 144 1979 OIL SPILL CONFERENCE

ually by a great number of soldiers. More than seven thousand men When all the visible oil had been removed, it was still necessary to worked along the coast picking up the seaweed and putting it in plastic try to clean the sand, the spoiled depth of which sometimes reached bags. These bags were then collected on the beaches by farmers' trac- fifty centimeters (20 inches) with two or three oil layers sometimes tors and trucks, and transported to dump sites in Brest and Tregastel. deeply buried in the beaches where extensive sand transfer had In some sections, front end loaders and trucks were used (Figure 6). occurred during each gale. Ploughing and harrowing the beaches to allow the sea to clean the sand were the most commonly used tech- niques. We tried in some places to mix with the oily sand a mixture of talc, water, and dispersant to make removal of the oil by the sea easier. Some attempts were also made to increase the natural biode- gradation of oil. Both artificial (soluble and insoluble) fertilizers and bacterial cultures were poured on the oily sand before harrowing. In some places where we had observed up to two kilograms per square meter (0.4 pounds per square foot) of oil mixed with the sand down to thirty centimeters depth (12 inches), much smaller oil quanti- ties were observed three months later (about 0.5 kg/m2, or 0.1 pound per square foot). Our laboratory analysis is not yet sufficient to be sure to what extent the applied agent assisted or accelerated the na- tural biodegradation process. In the case of shingle strands, cleanup was very difficult because the oil passing between the shingles is able to reach great depths where large amounts of it can be trapped. As these strands were not popular tourist areas, cleanup operations were conducted only in the places where the oil picked up by the sea would be able to re-pollute some cleaned beaches. In these areas, polluted shingles were pushed down by bulldozers to the level of low tide; the waves and the incoming tide then carried these shingles back toward the upper parts of the strand. The shingles actually scrubbed each other clean—or almost Figure 6. Front-end loaders removing oily sea weed from a clean—from the wave action after a few high tides. Some attempts beach near Portsall were also made to pour pure dispersant on the shingles and to let the tide then clean them up; in this case, however, the dispersant appeared to be non-effective. The thick layer of mousse which still spoiled the upper part of the In all the small harbors, from Porspoder to Brehat Island, piers and beaches was picked up, mostly by men with shovels and buckets, but slips were covered with oil, as were the beautiful pink granite rocks of sometimes with public works loaders (Figure 7). Some attempts were Tregastel and Perros-Guirrec. The economic impetus to cleanup the made to thicken the mousse layers by scraping and gathering the oil piers and slips for use by fishermen was as great as the economic im- with public works scrapers or by twenty meter long booms dragged petus for cleaning the rocks of Perros-Guirrec or Plougasnou for the over the sand behind two tractors. These two techniques required flat tourist business. At all these places, extensive and meticulous cleanup beaches without any rocks, but were sometimes very helpful, specially was required. Fire hoses were used first. When this low pressure (7 when the wind pushed mousse into a corner of the boom. This bars, or 100 psi) equipment proved no longer effective, very high pres- mousse, when collected, contained a large amount of sand and was sure pumps (up to a limit of 400 bar or 1,000 bar, 5,800 psi to 13,000 kept, as was the oily seaweed, for further treatment. psi) were tried. These pumps had to be washed with fresh water each evening; in addition, use of this very high pressure was dangerous to the concrete parts of the structures, as well as to personnel. It was also expensive. Such processes were effective between the sixth and ninth weeks after the Amoco Cadiz grounding. After that, high pressure, hot water pumps working at a pressure of 140 bars (2,000 psi) with a rela- tively low flow rate of fresh water heated to between 80° to 140°C, were used with very good efficiency under all circumstances for the next few months (Figure 8). One man could clean more than 500 square meters (about 5,000 square feet) of rocks per day at the begin-

Figure 7. Hand removal of emulsified oil covering a beach

Some attempts were also made to use a sorbent to facilitate oil re- moval. Sawdust, straw peat, shredded paper, leather powder, and rubber powder were all tested. The light products flew away with the wind and were not at all efficient. It had been hoped the rubber and leather powders would be efficient and rather cheap, but a difficulty arose in that the powder needs some energy for mixing with the oil or mousse. In such a case, only manual agitation was available. As a re- sult, these procedures needed more energy than appeared feasible and Figure 8. Cleaning of rocks using high pressure, hot water accordingly were not used. hosing technique CASE HISTORY — THE AMOCO CADIZ 145

ning, but four months later, the efficiency was much lower, 20-50 2. To burn the debris in a large incineration plant to be built at Brest square meters per day. To prevent the oil from sticking again on the for this purpose and afterwards used by the Brest tanker ship de- cleaned-up rocks after the next high tide, the best method appeared to ballasting facilities at a few percent of its capacity—This be to add a small amount of dispersant to the water. The initial approach would have been much more expensive than the first amount (0.5 percent) of dispersant was not adequate to provide a and was not further studied. steady dispersion so the percentage was increased to 3 percent. 3. To treat the debris with chemical products in order to separate oil Where piers had been built out over sandy fíat ground, trenches and debris, then washing in devices specially-built or adapted for were dug in which the oil and water collected and from which the oil this purpose—This approach also appeared to be much more ex- could be pumped or absorbed by rubber powder for manual collec- pensive than the first approach. tion. However, this method proved slow, tedious, and inefficient. For Many difficulties have also been encountered in dealing with the the most part, dispersants were added to the water of the washing products found at the bottom of the "honey-wagons", the trucks, and machines (0.5 percent of concentrated dispersant in the water), the the ships. These residues are actually deposits that cannot yet be de- flow rate of which was about one thousand liters per hour (about 4.5 emulsified. Carriage of these residues also appears quite impossible U.S. gallons per minute). Use of dispersants in this case had to be con- for they are too viscous for pumping, yet not solid enough for free- sidered from the ecological point of view as well. standing in trucks. Because some of the rivers contained beds, marshes had to be cleaned. Cleaning the marshes, especially the He Grande marsh near Tregastel which had been covered with oil, was done by removing the oil manually; in fact, in some places, 30 centimeter (one foot) thick layers of mousse could be found. To do this, some paths had to be built across the marsh. Unfortunately, we cannot yet say how long it will be before the marsh looks like it did before the spill. As of October, 1978 the cleaning of the river banks is not complete and certainly is the most difficult problem yet to be solved. These banks are covered either with shingles or with mud. In both cases, they are rather soft, yet we are obliged to wash off the oil without burying it. Because of the ecological vulnerability of the rivers, use of dis- persants was not recommended. Instead, the oil was pushed down by low pressure water streams to the river for collection there. The oil in this instance is collected by a particular device, "Egmolap", devel- oped by a local firm. This device is able to collect any kind of floating matter in a sheltered area. In the muddy areas, which are often very flat and impossible to walk on, it is also almost impossible to wash out the small amounts of oil which appear on the surface of the mud. In fact, after six months it

is still possible to find oil one foot inside the mud due to later deposi- tions of oily mud. The best way to remove this oil so far is to wash the muddy areas drastically with a mixture of water and a very light inorganic sorbent. The polluted sorbent is then collected in the same Figure 9. Artificial pond near Brest Merchant Harbor being way as noted above. This sedimentation of oil by muddy particles is filled with the more liquid portions of oily waste from the Am- one of the major continuing problems because the process helps oco Cadiz spill maintain a high concentration of oil on the bottom of the rivers.

Disposal and treatment of the oily debris

As the result of all the cleanup operations, many thousands of tons of oily debris accumulated in Brest and in Tregastel. The disposal of this debris created yet another set of problems. The oily debris was di- vided into different storage areas according to its fluidity and its oil content. The most liquid products (some 30,000 tons) were stored in five large ponds dug near the merchant harbor of Brest in an embank- ment area. These pools were lined with very strong plastic sheets care- fully welded together to be oil tight (Figure 9). After a very careful study, it now seems that the best treatment for these products would be water plus dispersant washing to separate the oil, which could then be sent to the local refining plant. The washed, but still oily, debris would then be treated along with the second type of debris which pri- marily originated from the beaches and rocky areas, and contains sand and seaweed, with less than 5 percent oil. About one hundred tons of this material was stored in Brest and forty thousand tons in Tregastel. The quickest and cheapest way to dispose of this debris has been to mix it with quick-lime, thereby creating a sort of material that, Figure 10. Artificial pond near Brest Merchant Harbor filled although it cannot be used for public works fill, still remains quite with oily debris which has been treated with quick lime inert (Figure 10). However, this stability may not be a long-term one; nevertheless, the oil concentration in the material is low enough to be absorbed by the quick-lime. Other possibilities, because of the huge amount of debris which had to be eliminated, have been considered. These are: 1. To burn the debris in the place where it was collected as many per- Conclusion sons actually suggested—However, this would have required a great number of small incinerators, each having a poor efficiency Those administrators facing this very large oil spill have tried to and using a large quantity of fuel oil to burn a small quantity of find quickly the most efficient solutions. Because of the size of the oily debris containing considerable water. spill, it provided the opportunity to try a number of various solutions 146 1979 OIL SPILL CONFERENCE in a great variety of circumstances. Some of the solutions appeared to Bibliography be the right way to go, but that was not necessarily so in other cases. The lay of the land was such that we were obliged to find effective 1. Commission of Enquiry of the Senate (Commission d'Enquete du ways to remove the oil and to clean the beaches without using heavy Senat), 1978. La catastrophe de ΓAmoco Cadiz. Official Journal equipment or a great amount of dispersant. of the Senate. Paris, France Some problems still remain as a result of the initial cleanup effort. 2. Ministry of Transport (Ministere des Transports, Direction des For example, the problem of oily waste disposal is complicated by the Ports et de la Navigation Maritime), 1978. Internal reports large amount of material having a very low content of oily debris. To regarding A moco Cadiz oil spill cleanup. Paris, France avoid this problem in the future, we are also studying all sorts of sce- 3. National Oceanic and Atmospheric Administration, Environ- narios in order to find the best way to fight a major new oil spill mental Research Laboratories, 1978. The Amoco Cadiz Oil should one ever occur again. Spill—A Preliminary Scientific Report. U.S. Superintendent of Documents, Washington, D.C., 20402 Acknowledgment

The authors are very grateful to their friends of the United States Coast Guard for courtesies rendered, especially to Lieutenant (jg) Robert Douville, USCG, who very kindly corrected this paper.